Grounded AIJosh Longenecker on evaluating AI systems, decision models, and other things worth measuring.https://groundedai.tech/Half the latency, 61% cheaper: cascading a decision model in front of an LLMhttps://groundedai.tech/blog/cascade-evaluator-tool-calls/https://groundedai.tech/blog/cascade-evaluator-tool-calls/CascadeEvaluator lets TypeSafe’s Jev settle the decisions it is sure of and passes the rest to any LLM judge you pick. Against Claude Sonnet 4.6 on 200 real agent turns, it was 51% faster and 61% cheaper, with no loss in accuracy.Thu, 08 Oct 2026 00:00:00 GMTevaluationdecision modelsagentsgrounded-ai